The proposed project is designed to advance our understanding of the development and mechanism of categorization - a fundamental component of human intelligence. The proposed research is based on the hypothesis of multiple mechanisms sub-serving category learning: (a) an early developing mechanism that is based on distributed attention and learning of within-category statistics and (b) later developing mechanism that is based on selective attention to category-relevant information. To test these hypotheses, we will conduct a series of experiments with 8-20 month-old infants, 3-8 year-old children, and adults. The proposed project has the following Specific Aims. Specific Aim 1 is to examine developmental changes in category learning in infancy and early childhood. To achieve this aim, we will create category structures that can be learned by either distributed or selective attention, with each pattern of learning resulting in different category representation. By focusin on attention during category learning and on patterns of generalization, we will infer how categories are learned and represented. By comparing category learning and representation across ages under the same task and by manipulating the category structure we will infer how category learning changes with development. Specific Aim 2 is to link the development of selective attention with the development of category learning. To achieve this aim will measure selective attention independently of category learning and will link variability in selective attention with variability of learning category structures that require focused attention. The proposed project will elucidate the development and mechanism of categorization - one of the most critical components of cognitive development. In addition to contributing to our understanding of typical cognitive development, these answers may also contribute to the early diagnostics and treatment of several intellectual disorders, such as autism and Williams syndrome.